POLICY PROPOSAL 2018-03 | FEBRUARY 2018

The Importance of Strong Labor Demand

Jared Bernstein MISSION STATEMENT

The Hamilton Project seeks to advance America’s promise of opportunity, prosperity, and growth.

We believe that today’s increasingly competitive global economy demands public policy ideas commensurate with the challenges of the 21st Century. The Project’s economic strategy reflects a judgment that long-term prosperity is best achieved by fostering economic growth and broad participation in that growth, by enhancing individual economic security, and by embracing a role for effective government in making needed public investments.

Our strategy calls for combining public investment, a secure social safety net, and fiscal discipline. In that framework, the Project puts forward innovative proposals from leading economic thinkers — based on credible evidence and experience, not ideology or doctrine — to introduce new and effective policy options into the national debate.

The Project is named after Alexander Hamilton, the nation’s first Treasury Secretary, who laid the foundation for the modern American economy. Hamilton stood for sound fiscal policy, believed that broad-based opportunity for advancement would drive American economic growth, and recognized that “prudent aids and encouragements on the part of government” are necessary to enhance and guide market forces. The guiding principles of the Project remain consistent with these views. IThe Importance of Strong Labor Demand

Jared Bernstein Center on Budget and Policy Priorities

FEBRUARY 2018

This policy proposal is a proposal from the author(s). As emphasized in The Hamilton Project’s original strategy paper, the Project was designed in part to provide a forum for leading thinkers across the nation to put forward innovative and potentially important economic policy ideas that share the Project’s broad goals of promoting economic growth, broad-based participation in growth, and economic security. The author(s) are invited to express their own ideas in policy papers, whether or not the Project’s staff or advisory council agrees with the specific proposals. This policy paper is offered in that spirit.

The Hamilton Project • Brookings 1 Abstract

By conventional measures, the U.S. job market has suffered some degree of slack for about 70 percent of the time since 1980. The absence of persistent, strong labor market demand has a significant negative impact on wages and incomes, with these costs falling disproportionately on the least advantaged. In this paper, I offer a four-part proposal to increase labor demand along with earnings and employment opportunities: (1) reform our monetary policy framework to accommodate more monetary stimulus and reduce the risk of hitting the zero lower bound, (2) develop a Full Employment Fund to reduce labor market slack, (3) support direct job creation programs to boost labor demand, and (4) design international trade policies to safeguard aggregate demand and mitigate the negative effects of trade deficits.

2 The Importance of Strong Labor Demand Table of Contents

ABSTRACT 2

INTRODUCTION 4

THE CHALLENGE 6

A NEW APPROACH 11

QUESTIONS AND CONCERNS 18

CONCLUSION 19

AUTHOR AND ACKNOWLEDGEMENTS 20

ENDNOTES 21

REFERENCES 22

The Hamilton Project • Brookings 3 Introduction

t is a remarkable fact that since 1980, by one conventional opportunities have long been elusive. But for economists relying measure, there has been slack in the labor market far more on models that assume full employment, as many models do, the often than not. That is, there has often been insufficient fact that the U.S. economy has been at full employment less than Idemand for labor, putting downward pressure on job a third of the time since 1980 is an awfully inconvenient truth. opportunities and wage growth. It is also the case that many of the troubling trends in our Figure 1 shows the difference between the unemployment rate economy, including wage and income stagnation, along with and a frequently used estimate by the Congressional Budget the rise of inequality, occurred largely after 1980. Of course, the Office (CBO) of the so-called natural rate of unemployment, or absence of full employment is only one factor in those outcomes. the rate economists believe to be the lowest jobless rate consistent Expanded trade and technological advances have contributed with stable inflation. Though this paper critiques this concept to slower wage and employment growth for certain groups of of a reliably identifiable natural rate, by this broadly accepted workers. In addition, the loss of union power, the erosion of measure, the U.S. job market has been slack about 70 percent labor standards (e.g., minimum wage levels and the overtime of the quarters since 1980, compared to just about a third of the salary threshold), and corporate consolidation and greater quarters from 1949 to 1980. market power of large firms have all tilted the playing field against less-advantaged workers. These factors help to explain This fact of persistent slack might not be viewed as remarkable the set of adverse wage and income outcomes for workers over by many Americans stuck in places where gainful employment the past few decades.

FIGURE 1. Labor Market Slack, 1949–2017

6

4

Positive slack 2

0

Slack (percentage points) (percentage Slack –2

Negative slack –4 1953 1957 1961 1965 1969 1973 1977 1981 1985 1989 1993 1997 2001 2005 2009 2013 2017

Source: Current Population Survey 1949–2017; CBO 2017. Note: Labor market slack is defined as the difference between the actual unemployment rate and the natural rate of unemployment: a positive slack value indicates elevated unemployment. 2017 values are based on the first three quarters of the year.

4 The Importance of Strong Labor Demand But weak aggregate demand—the total demand for goods and Moreover, unlike many of the factors that dampen wage levels services throughout the economy—is an especially pervasive and growth, including eroded labor standards, arguments problem with unique characteristics. By definition, it suggests in favor of strong aggregate demand do not tend to provoke resource underutilization, which implies some degree of market partisan rancor; in principle, policymakers generally agree failure, thus warranting a policy response. Similar to falling on the need for strong demand. That said, policymakers unionization, weak demand erodes the ability of many in the have not yet taken adequate steps to keep the economy at workforce to bargain for higher compensation. Even in the full employment, as is evident from figure 1. Clearly, the absence of unions, strong demand leads employers to bid up problem of inadequate demand is not deemed sufficiently their wage offers to get and keep the workers they need if they are urgent by enough policymakers, perhaps because, as I show to meet consumer demand. In slack labor markets, such wage in the section on labor market tightness and wage growth, its pressures abate. downsides are concentrated among the least well-off.

Persistent slack has also been shown to lead to lasting (as Precisely what steps would ameliorate the problem of excessive opposed to temporary) negative effects on the supply side of labor market slack is the subject of much debate. Because there the labor market and the broader economy. Even temporary is no consensus about how to solve the slack problem, partisans shocks can cause permanent damage if workers’ skills erode often argue for their favorite solutions—tax cuts recommended or if spells of long-term unemployment lead them to give up by conservatives or infrastructure build-outs suggested by and permanently leave the job market. A recent, rigorous look progressives—with insufficient evidence and economic rationale. at these effects in the labor market finds that workers in areas To improve this discussion, I first examine the relevant evidence with relatively large unemployment shocks during the Great and economic theory, then propose policies to boost aggregate Recession had significantly lower employment and earnings demand that are rooted in that assessment. years later (in 2015), relative to similar workers in places with milder upticks in unemployment (Yagan 2017). These impacts I propose a four-part policy response. First, the monetary were particularly damaging for lower-wage workers, presaging policy framework should be reformed to reduce the risk of results shown later in this paper on the relative impact of slack hitting the zero lower bound (ZLB) and to ensure that the at different wage levels. central bank has the ability to support the economy during a downturn. Second, we must expand our thinking about fiscal Other research shows the long-run impact of demand shortfalls policy and aggregate demand beyond recession-fighting to on potential and actual gross domestic product (GDP), though encompass sustained fiscal policy during weak expansions. I economists remain uncertain how much of that loss is truly therefore propose a mandatory Full Employment Fund (FEF) attributable to persistently weak demand. DeLong, Summers, that expands and contracts with need. Third, as a complement and Ball (2014) argue that much of the post-2007 gap between to this fund, I propose measures providing for direct job earlier and later vintages of CBO’s estimates of potential creation. Finally, I note that in the presence of the ZLB, GDP—in other words, the decline in CBO’s estimate of persistent trade deficits can constitute a drag on aggregate potential GDP in a given year—can be attributed to transitory demand, and I propose policies to both restore lost demand shocks becoming permanent. In the second quarter of 2017, and reduce the trade deficits themselves. that difference amounted to just over $2 trillion, which is the difference between the 2007 projection of potential GDP in This proposal begins with an analysis of the historical extent 2017Q2 and the 2017 calculation of potential GDP in 2017Q2. of economic slack—the persistent absence of strong aggregate It amounts to a loss of about $6,500 per capita. demand—and then turns to an analysis of the impact of economic slack on wages and incomes. I then develop a policy Even if only a part of that amount is attributable to the agenda intended to significantly raise the amount of time impact of persistent slack, weak aggregate demand is clearly a during which the U.S. labor market is at full employment. costly problem, suggesting the need for policies to address it.

The Hamilton Project • Brookings 5 The Challenge

A BRIEF HISTORY OF SLACK AND OUTPUT GAPS IN is, if the natural rate is lower than typical estimates suggest, the THE U.S. ECONOMY actual unemployment rate minus the natural rate would yield Any efforts to identify the extent of slack quickly run into larger slack estimates than shown in figure 1. measurement challenges. Estimating slack requires either a In fact, the difficulty in finding a trustworthy measure of the calculation of the natural rate of unemployment or the output natural rate of unemployment is evident in figure 3, which gap between actual and potential GDP, in which case we are shows the evolution over time of both point estimates of the invoking variables that we cannot directly observe (see box 1). natural rate and confidence intervals surrounding it (note Moreover, both of these capacity measures have come under that these estimates differ from those shown in figure 2). Over scrutiny in recent years, leading to portentous questions about the past 20 years the natural rate has moved around a bit, but their value as policy guideposts. more importantly, our ability to estimate it with the degree Figure 2, for example, plots the estimate of the Federal of accuracy necessary for policymakers has collapsed. This Reserve’s (Fed’s) natural unemployment rate against actual decreasing precision follows from the diminished correlation unemployment, wage growth, and both actual and targeted between unemployment and inflation, which is the traditional inflation rates. As the unemployment rate fell sharply from 10 basis for calculating the full employment rate. As such, the percent to almost 4 percent (the January 2018 unemployment declining precision reflects the dynamics shown in figure 2, rate of 4.1 percent is the lowest since December 2000), inflation with inflation becoming less responsive to changes in slack. has not accelerated at all, and nominal wage growth increased Figure 4 compares a more comprehensive slack measure, only slightly. Such dynamics suggest various possibilities, derived by economist Andy Levin (2014). His gap measure including a low responsiveness of inflation to unemployment comprises three equally weighted parts: the gap between the and/or that there is more slack in the labor market than unemployment rate and the natural rate, the gap between the suggested by the unemployment rate. If that is the case, then labor force participation rate and its expected value at full the slack suggested in figure 1 could be underestimated. That employment (as per CBO), and the hours-weighted share of

BOX 1. Measuring Slack in the Labor Market

Two estimated variables are typically used as benchmarks for calculation of labor market slack: first, the so-called natural rate of unemployment; and second, potential GDP. These variables are not directly observed, but must be inferred from other data in the context of a particular economic model. The natural rate of unemployment is the hypothetical lowest jobless rate at which price growth (inflation) would remain low and stable. If actual unemployment stays below this level for a while, we would expect inflation to accelerate. Conversely, when actual unemployment is above the natural rate, we would expect inflation to remain subdued and workers to suffer weak labor demand. Potential GDP, also referred to as potential output, is the level of economic output that is possible at a given time if labor and capital are fully utilized. When actual GDP falls below potential—that is, when there is slack in the economy—not all available resources are being utilized. Both potential GDP and the natural rate of unemployment are unobservable variables that must be inferred from other, observable relations, such as the correlation between inflation and unemployment. Because these correlations change over time and across place, estimates of potential GDP and the natural rate of unemployment are subject to considerable uncertainty.

6 The Importance of Strong Labor Demand the workforce that is underemployed (i.e., involuntary part- to expectations, and the share of underemployed workers was time workers). Note that since around 1980 the Levin gap notably elevated in this business cycle relative to past cycles. is larger than the standard gap measure; this difference was relatively large during the Great Recession and subsequent Potential GDP—the level of output at full resource utilization— slow recovery. This was driven by both additional factors in and the output gap between real and potential GDP are also the Levin measure: labor force participation was low relative estimated with uncertainty (see box 1). Turning to the output gap,

FIGURE 2. Unemployment, Wage Growth, and Inflation, 2007–17

10

9

8

7

6

5 Natural rate of

Percent unemployment 4 Actual unemployment 3 Wage growth 2 In ation target PCE core 1 in ation

0 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017

Source: Bureau of Economic Analysis (BEA) 2017; Current Population Survey 2007–17; Federal Open Market Committee (FOMC) n.d.; author’s calculations. Note: PCE is the personal consumption expenditures price index. 2017 values do not include December. Data for natural rate of unemployment are only shown starting in 2009.

FIGURE 3. Estimates of the Natural Unemployment Rate, 1978–2015

8

7

6

5 Natural rate of 4 unemployment

3

2

Unemployment rate (percent) rate Unemployment 1

0 1980 1985 1990 1995 2000 2005 2010 2015

Source: Council of Economic Advisers 2016. Note: Shaded gray areas indicate 50 percent confidence intervals.

The Hamilton Project • Brookings 7 figure 5 shows three quite different estimates of potential GDP economy; and total factor productivity (innovation), as opposed since just before the Great Recession, along with actual GDP. to temporary demand shocks. Recent research by Coibion, Two of the lines track CBO estimates of potential, derived from Gorodnichenko, and Ulate (2017) finds that such measures often a combination of trend extraction and a bottom-up aggregation conflate supply and demand shocks.1 of estimates of production factors and productivity at full employment. The critical aspect of these estimates is that they The implications of these figures are at least twofold. First, and are designed to capture lasting, structural changes in supply-side most importantly, the U.S. labor market has been slack more variables, including the stock of human and physical capital in the often than not, as shown, for example, by the comparison of the

FIGURE 4. Labor Market Slack by Measurement Method, 1960–2017

8

6

4

2 Levin gap 0 Traditional slack measure (u – u*) Slack (percentage points) (percentage Slack –2

–4 1963 1969 1975 1981 1987 1993 1999 2005 2011 2017

Source: Andrew Levin (personal communication); Current Population Survey 1960–2017; CBO 2017. Note: 2017 values are based on the first three quarters of the year. The traditional slack measure is the unemployment rate minus the natural rate of unemployment.

FIGURE 5. Actual GDP vs. Potential GDP, 2007–17

19 Coibion et al. 2017 January 2007 CBO 18

June 2017 CBO 17 Actual GDP

16

15 GDP (trillions of 2009 dollars)

14 2007 2009 2011 2013 2015 2017

Source: BEA 2007–17; Coibion, Gorodnichenko, and Ulate 2017; CBO 2007, 2017; author’s calculations. Note: 2017 values are based on the first two quarters of the year. January 2007 and June 2017 CBO numbers are based on 2007 and 2017 Congressional Budget Office estimates of potential GDP; Coibion et al. 2017 shows potential GDP as estimated in Coibion, Gorodnichenko, and Ulate (2017) using an econometric tech- nique developed by Blanchard and Quah (1989).

8 The Importance of Strong Labor Demand actual unemployment rate to CBO’s estimate of the natural rate. The first set of results (figure 6) is derived from a state- Such persistent slack puts downward pressure on wage growth, level analysis of how wages respond to changes in either both nominal and real, which motivates a key theme of this unemployment rates or employment-to-population ratios.2 As chapter: implementing aggregate demand-side policies to get to expected, increases in employment lead to increases in wages, and stay at full employment is instrumental in boosting wage and increases in the unemployment rate lead to decreases and income growth, especially for less-advantaged or lower- in wages. Notably, the impact is much larger on low-wage wage workers. workers, and, in fact, for high-wage workers slack and wages appear to be unrelated. Second, economists cannot, within a policy-relevant confidence interval (i.e., an interval that could reliably drive policy The magnitude of the impacts is economically meaningful. decisions), accurately calculate the extent of slack in the job For example, as the U.S. job market moved to full employment market or broader economy. Absent clear signs of utilization during the 1990s, the jobless rate fell from 7.5 percent in 1992 constraints, and weighing both the macro and micro costs to 4 percent in 2000. Over that period the 20th percentile of of weak demand against the risks of inflation, policymakers real wages grew 10 percent and median real wages grew 4 seeking to address wage stagnation, high levels of inequality, percent, implying that about 70 percent of each increase is and weak worker bargaining power would be advised to associated with the unemployment decline. aggressively apply the policies discussed later in this proposal. This relationship between slack and hourly pay has long LABOR MARKET TIGHTNESS AND WAGE GROWTH been understood in economics, particularly with respect to nominal pay. In addition, wage curve analysis has uncovered The first part of this section establishes that slack has been relationships like those shown in figure 6, all implying common in the U.S. labor market. This section shows the substantial wage gains when slack is lower. But there is another impact of slack on wages and incomes, with a focus on the favorable effect of diminished slack, one that can be even more distributional impacts. These findings reveal economically and dramatic in terms of its impacts on the income of working statistically significant negative impacts of slack on real wages families: the way low levels of slack can increase labor supply. and incomes. Moreover, these costs fall disproportionately on the least advantaged; in fact, correlations between slack and For working families, annual income can be simply defined high wages tend to be statistically indistinguishable from zero. as earnings plus nonlabor income. The earnings term can be

FIGURE 6. Wage Differences Associated with Increases in Unemployment Rate and Employment-to- Population Ratio

8 Employment-to-population ratio 6

4

2

0

−2

−4

Change in real hourly wage (percent) hourly wage in real Change Unemployment rate −6 10 20 30 40 50 60 70 80 90

Wage percentile

Source: Current Population Survey 1979–2015; Economic Policy Institute 2017; author’s calculations. Note: Bars in the charts show the impact on wages of a one-standard-deviation increase of a labor utilization variable over the 1979–2015 period. Hollow bars indicate coefficients that are not statistically significant at the 5 percent level.

The Hamilton Project • Brookings 9 FIGURE 7. Simulated Real Median Household Income by Rate of Jobs Recovery, 2007–16

62

60 Baseline 58

56 Slower recovery 54 No recovery 52 Median household income (thousands of 2016 dollars) 50

48 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016

Source: Current Population Survey 2007–17; CBO 2017; author’s calculations. Note: “Baseline” shows the actual level of median household income, “Slower recovery” shows a simulation in which employment is assumed to grow half as quickly and unemployment fall half as fast as actually occurred, and “No recovery” shows household income with no post-2010 improvement in employment.

usefully decomposed as follows: Annual income = earnings about 80 percent of the variance in nominal median household per hour × hours per week × number of weeks + annual income.3 nonlabor income. Using this model, I simulate the evolution of real median Slack does not matter only for hourly earnings: significant household income under the assumption of no post-2010 relationships similar to those shown in figure 6 exist between improvement in employment. I also conduct a similar slack and both hours per week and the number of weeks simulation in which employment is assumed to grow half worked. Bernstein, Spielberg, and Bentele (forthcoming) as quickly and unemployment fall half as fast as actually examine these relationships for low-wage workers over the occurred (shown in figure 7). Even though I allow hourly 1979–2015 period, focusing on the role of stronger demand wages to grow exactly as they did over 2007–16, real median (low state-level unemployment) in generating higher earnings incomes either fall or stagnate under the two simulations, and incomes through increased labor supply. For all low-wage revealing the importance for middle-class incomes of having (bottom quintile) workers, the impact of falling unemployment more work. That is, much of the recent improvement in real on labor supply raised annual earnings by about 20 percent. median household income has come not from wage gains, but For single mothers, lower unemployment raised earnings from increases in hours and employment rates. through the labor supply channel by 54 percent; for African Americans, 43 percent (Bernstein, Spielberg, and Bentele To be sure, more work at stagnant hourly earnings is costly forthcoming). to families in terms of reduced time for leisure and family responsibilities. Given the real hourly wage stagnation for In other words, while stronger labor demand puts upward low- and middle-wage men in the 1980s and both men and pressure on wages, it also adds to annual earnings through women in the 2000s, to the extent that incomes rose during increased labor supply. Another way to see this is to build a those periods, those increased incomes were largely a result of time-series model of median income growth as a function of more work. Hourly wage stagnation is, in other words, far from inflation, employment, hourly wages, and slack (measured costless. But the record also shows that strong labor demand as the gap between the unemployment rate and the natural raises incomes through increases in both employment and rate of unemployment). A simple model as described explains hours worked.

10 The Importance of Strong Labor Demand A New Approach

he preceding analysis shows that, by various commonly MONETARY POLICY: TIME TO TRY SOMETHING NEW used metrics, the economy does not quickly return to Monetary policy is carried out by the U.S. Federal Reserve, full employment after recessions, labor market slack which has a well-known dual mandate of maintaining full Tis common, and this slack is costly, especially to less well-off employment at stable prices. Thus, the work of the Fed is at families. For this reason, we need a policy agenda that will the heart of maintaining strong aggregate demand. My focus squeeze more slack out of the U.S. job market. The rest of this is, of course, on the employment side of the Fed’s mandate, proposal explores such an agenda. but understanding the role of price pressures in pursuit of tight labor markets is critical to achieving and maintaining These proposals fall into four general categories: monetary, full employment. As I argue next, countercyclical fiscal policy fiscal, direct job creation, and international trade/finance must of course be part of the response to temporary demand policies (though direct job creation is a specific application contractions, but the first line of defense is typically monetary of fiscal policy). Since the goal of this agenda is to not only easing by the central bank. get to, but also to stay at full employment, I also consider financial regulatory policies to be highly germane because, in The Fed faces two significant challenges in terms of recent decades, financial bubbles have been a potent enemy of maintaining strong aggregate demand. First, as suggested maintaining tight labor markets. However, in the interest of in the preceding two sections, the Fed does not have reliable brevity I say little about these issues here. Also, while the focus guideposts as to what constitutes full employment or potential is mostly on demand-side policies, I envision but do not discuss GDP. If the Fed sets the natural rate too high or potential a role for training and apprenticeships within direct job creation GDP too low (as Coibion, Gorodnichenko, and Ulate [2017] programs. Updating and maintaining strong labor standards– suggest to be the case), that action creates a risk that it will including minimum wages, labor unions, and overtime pay wield interest rate policy in ways that keep the economy from rules—are key to a progressive wage agenda, but my focus here achieving sufficient aggregate demand to tap the benefits for is more narrowly on policies to boost aggregate demand. less-advantaged workers shown in the previous section.

I propose that the following national policies be enacted to The second challenge for the Fed is that when short-term reduce labor market slack and raise labor demand: nominal interest rates have been reduced to zero, the central bank can no longer stimulate the economy through its most • Monetary policy: Change inflation targeting at the Federal powerful weapon: lowering the interest rate it controls, Reserve to both accommodate more monetary stimulus and thereby reducing the cost of borrowing and investing. While reduce the risk of encountering the zero lower bound (ZLB) to interest rates. ZLB risk is at the core of all the proposals: the trade some central banks have reduced interest rates slightly below deficit, for example, poses a greater threat to labor demand when zero, the U.S. Fed has heretofore not gone this route and Fed interest rates are near zero. officials have not suggested that this is a tool they would readily use (Irwin 2016). Economists discuss this problem • Fiscal policy: Develop an automatic Full Employment Fund (FEF) of the effective lower bound on the policy interest rate as the that expands and contracts with changes in the business cycle. zero lower bound (ZLB). While lowering the rate it controls is • Direct job creation: Design the FEF so it will support direct not the sole tool in the Fed’s toolbox, it is widely agreed that job creation programs, from subsidized employment to public hitting the ZLB is a serious constraint on generating more service jobs. aggregate demand. • International policies: Implement policies to ensure that changes in global demand are not a drag on aggregate demand within the While the focus of this proposal is on longer-term weakness United States, especially when there is already persistent slack in on the economy’s demand side, current events are instructive the U.S. labor market. of the longer-term challenge. Look back at figure 2. Clearly, unemployment has fallen below the Fed’s natural rate (4.6

The Hamilton Project • Brookings 11 BOX 2. Why Is the Zero Lower Bound Important?

The main policy tool of most central banks is to set an overnight borrowing rate that banks use to borrow and lend to one another. By adjusting this benchmark rate, central banks have impacts on a wide range of interest rates that help determine economic activity, such as car and mortgage loans. If there is a large enough negative shock to the economy, the central bank may reduce that rate to zero. In that case, it can provide no additional stimulus to the economy via rate cuts. Given the current structure of our economy and financial system, zero becomes a boundary: if interest rates were substantially negative, depositors could remove money from banking systems and hold cash instead. If a shock is large enough that zero is not a sufficiently low interest rate to restore demand in the economy and move the economy toward full employment, the economy is said to be stuck at the zero lower bound (ZLB). There are other tools the central bank can use to influence the economy even if it is at the ZLB. For example, it can make promises regarding how long it will keep rates low to try to lower long-term interest rates. Alternatively, it could buy long-term government bonds or mortgage-backed securities to try to directly change key interest rates. The Fed has used a variety of tools in the past decade, ranging from direct buying (often referred to as quantitative easing) to making commitments about future rates (i.e., forward guidance). The impact of these tools is widely debated, but most economists agree that central banks’ ability to provide monetary stimulus is constrained when their policy rate is at the ZLB.

percent as per its latest projections), and yet core inflation has as savings have outpaced investment. (A contributor is the decelerated (FOMC 2017). Nominal wage pressures have also savings glut problem associated with countries with persistent remained subdued. trade surpluses—explored in the final section of A New Approach.) This has led to arguments against preemptive tightening that could prevent the benefits of tight labor markets from These facts imply that hitting the ZLB, as occurred in the Great reaching many who have heretofore been left behind in this and Recession, is a greater risk going forward than it has been in the prior expansions. But even sympathetic members of the Fed, past. It is hard to overstate the downsides of this risk. Though including former Chair Janet Yellen (a strong advocate of full some banks have set rates below zero, the ZLB remains a employment), worried that the Fed could get behind the curve threatening constraint that could be increasingly worrisome if and that inflation would become de-anchored; such fears push the equilibrium interest rate remains historically low. By setting a the Fed toward raising rates. higher inflation target, equilibrium nominal interest rates would be higher, making it less likely that the central bank would reduce One way to ensure that the Federal Reserve uses policy in a interest rates to zero. way to maintain sufficient aggregate demand while addressing a number of changing macroeconomic realities and growing Targeting a higher inflation rate or level has other useful attributes. risks, particularly ZLB risk, is for the Fed to raise its inflation Particularly in a period like the present, with a tightening job target. Better yet, the Fed should shift to targeting the level market amid weak price pressures, a higher target would lead of a key variable, like the price index, nominal GDP, or the to a more patient Fed, one that would allow the benefits of full nation’s wage bill. While any such changes would be large and employment to be felt more broadly before it acted to slow the potentially disruptive, they could be helpful in more reliably economic expansion. sustaining aggregate demand. In a recent review of these issues, Binder and Rodrigue suggest One key impact of a higher inflation target would be to that “in terms of reaching full employment, price-level targeting provide the Fed more weaponry against ZLB risk, as well as may be more effective than inflation targeting.” They argue, for demand contractions. Extensive research finds that interest example, that a “central bank using price-level targeting would rates have declined structurally across advanced economies reduce the output gap more aggressively than a bank using in recent years, and many monetary economists, including inflation targeting, thus keeping employment more stable” those at the Fed, argue that the economy’s equilibrium interest (Binder and Rodrigue 2016, 12–13). In periods of weak price rate—the interest rate consistent with full employment—has growth, like the current one, this effect is mechanical in the fallen as well (Williams 2017). Some researchers, including following sense: Suppose, after some period of inflation below its Larry Summers as part of his reintroduction of what he calls target, inflation reverted up to its target rate. The Fed would wash the secular stagnation problem, argue that persistently weak its hands and declare its stimulative work to be complete. demand is partly responsible for the decline in interest rates,

12 The Importance of Strong Labor Demand But under a level target, the Fed would be committed to allowing around the level it has been for a long while. Both the prices to rise more quickly than the target rate, in order to close Bernanke and Yellen Feds worked very hard to convey this the gap between the actual level and the targeted level that message, because they reasonably view anchoring to be a key developed over the period of weak inflation. This is because the determinant of stable prices. Add this to the fact that the Fed level target, unlike the rate target, must make up for past misses. has been undershooting its price target for a number of years, This difference implies that under a credible, level-targeting and we must admit that convincing the public of a change in central bank, periods like the past few years create expectations the Fed’s inflation target will be very challenging. of faster inflation, which in turn produce expectations of lower real interest rates, and thus greater demand. A more deliberative approach would be to organize a process by which central bankers along with outside advisers and Former Fed chair Ben Bernanke (2017) agrees that level- stakeholders can explore these issues—both that of the ZLB targeting is preferable to targeting a higher rate, and argues the and unreliable macro-guideposts—in a climate that is not latter is too costly in that “it forces society to bear the costs of fraught with market and political risks.4 The Fed should set up higher inflation at all times, instead of only transitorily after a time-limited commission—say, a year-long process—tasked periods at the ZLB.” He proposes an interesting hybrid: keep with considering whether a change to its current framework the 2 percent inflation target in normal times, and switch to regarding inflation—its 2 percent target—is warranted, and, if temporary 2 percent price-level targeting when rates are at the so, recommend a different framework. ZLB. This creates a lower-for-longer rate regime by the Fed’s interest rate setters, because they must make up for persistent To maintain a substantively and politically contained process, misses on inflation. This would have been relevant to the most the commission should accept the premise of the dual recent business cycle, given that core PCE inflation has been mandate. Accepting that premise obviates legislative changes: below the 2 percent Fed target for much of the past decade. the commission should discuss tools, not goals (the results of the commission would thus be advisory to the Fed, and would Binder and Rodrigue—in addition to many others—argue not be legislatively mandated). Careful consideration should that targeting an economic aggregate like nominal GDP is an be taken to ensure representation by those with the most at even better idea for maintaining aggregate demand (Brookings stake from the persistent slack shown in the beginning of Institution 2018). After all, if the ultimate problem we are this proposal, such as advocacy groups for minority and low- trying to solve is inadequate income or wage growth, why income workers. The commission’s meetings, findings, and not directly target the level of those variables? Since nominal papers should be made public, which would help to prepare growth is real growth plus inflation, either slower real growth markets and the broader public for a regime switch, if that or slower inflation would induce looser monetary policy. Again, is what is forthcoming. To avoid political risks, this process these targets are especially attractive in periods of protracted should be run by the Fed itself, and not by Congress. However, weakness (like much of the current recovery), during which the to achieve political buy-in, staffers from committees that deal Fed would signal that its goal was not just to get back to some with monetary policy (e.g., the Senate Banking Committee target growth rate, but to make up for lost ground by surpassing and the House Financial Services Committee) should also that growth rate for as long as was necessary. participate in the process.

Recently, some Federal Reserve officials, including former Given that the most recent few economic expansions fell Chair Yellen (Glassman 2017), former Vice Chair Fischer victim to imploding asset bubbles, the Federal Reserve’s (Robb 2017), and San Francisco Fed president John Williams macroprudential role—its oversight of the banking system—is (Harrison 2017), have all signaled some interest in these ideas. also germane to this agenda. The key policy recommendation However, the statements and musings of influential central is to use regulatory, and not interest rate, policy to push back bankers are always amplified, and sometimes misinterpreted, on potential bubbles and underpriced risk. That is, if financial by markets and investors, making it difficult for the Fed to markets become too effervescent, it is important to employ explore innovative monetary policy ideas, and consequently regulatory interventions (e.g., rules that reduce leverage ratios), subjecting the bank to a massive status quo bias. rather than interest rate hikes, as countervailing measures. Former Chair Yellen (2014), along with macroeconomists Also, while academics often suggest that the Fed should adjust Blanchard and Summers (2017) have recently underscored the its inflation target, as if this was merely a technical issue, in the benefits of this approach, and Lars Svensson (2017) provides real world it is surely difficult to change market expectations. empirical evidence in support of it. People and markets appear to have firmly internalized the current target rate and thus have come to expect the Fed to anchor inflation at either 2 percent or—more realistically—

The Hamilton Project • Brookings 13 SUSTAINED FISCAL POLICY THROUGH A FULL than higher future debt-to-GDP ratios. In a recent paper, EMPLOYMENT FUND Ben Spielberg and I suggest various ways to make Keynesian In 2013, when the U.S. economy had already been expanding for stimulus more effective, including increasing the role of about four years, Fed chair Bernanke stressed the importance automatic stabilizers, such as the Supplemental Nutrition of countercyclical fiscal policy in his congressional testimony: Assistance Program (SNAP), extending the duration of “Although monetary policy is working to promote a more robust unemployment insurance benefits, and increasing state fiscal recovery, it cannot carry the entire burden of ensuring a speedier relief (Bernstein and Spielberg 2016). return to economic health. The economy’s performance both These findings are all particularly relevant to boosting over the near term and in the longer run will depend importantly aggregate demand during recessions. However, it is also on the course of fiscal policy” (Bernanke 2013). important for fiscal policy to squeeze out the residual slack In fact, especially in recessions and weak recoveries, during expansions, and the next section expands on this monetary and fiscal policy can interact to boost aggregate idea. The idea behind what economist Jason Furman (2016) demand. The Fed’s firepower is diminished in periods of low calls sustained fiscal policy is that the related phenomena of equilibrium interest rates, and recent research suggests that weak recoveries and low interest rates, specifically interest fiscal policy is particularly effective at the ZLB (Auerbach and rates below growth rates, create the need and opportunity Gorodnichenko 2017). Unfortunately, the challenges faced for policymakers to make demand-strengthening public by the Fed in raising demand at the ZLB were exacerbated investments in recoveries. Furman writes, “Sustained fiscal by austere fiscal policy from 2011 through 2015, when policy may be necessary because the global economic climate policymakers engaged in fiscal consolidation rather than may be showing symptoms of persistently inadequate demand the needed expansion. (This very dynamic was the reason dragging on growth and inflation” (11). for Bernanke’s quoted comment above.) Moreover, recent When the national economy is in recession, most—though not research suggests a relationship between austerity measures, all, of course—economists accept the role of temporary fiscal weaker growth rates (Blanchard and Leigh 2013; Shambaugh stimulus. The idea of sustained fiscal stimulus is that, even in 2012), and even long-run impacts of weak demand on supply recovery, there are places and groups of people that have been (Ball 2014; Summers 2014), suggesting a very steep cost to such consistently left behind, such that recessionary conditions fiscal policy mistakes (see also figure 7). With that context can prevail in some parts of the country even when national in mind, this section offers proposals designed to avoid the unemployment is low. damaging bouts of fiscal austerity that have contributed to persistent slack in the U.S. economy. In addition, as long as the economy’s growth rate surpasses the interest rate—as has long been the case in the United Fiscal policy—tax, transfer, and spending policy by States and even more so recently—debt servicing costs should governments—can play at least three roles in boosting and remain manageable (Kogan et al. 2015). We find this dynamic maintaining aggregate demand. The first is the well-known, not only in U.S. data, but also in most advanced economies though sometimes disparaged, Keynesian role, wherein (Furman 2016). government spending temporarily ramps up to offset a demand contraction. The second role, and the one most relevant to this In this way, insufficient aggregate demand creates the necessary paper, is the use of fiscal policy to offset excess slack in recoveries conditions for sustained fiscal stimulus. The lowered propensity characterized by weak aggregate demand. Third, through public for private investment, higher global savings, and, in the investment in physical and human capital, fiscal policy can boost U.S. case, capital inflows all combine to push interest rates the supply side of the economy, raising potential growth and below growth rates. This leaves us with weakened demand, generating more labor market opportunities. the pervasive absence of full employment, the potential for permanently damaging supply-side impacts, and low borrowing Following the Great Recession, research on both the U.S. costs. The obvious solution is sustained fiscal investments and European economies has strengthened the case against targeting the people and areas where demand is weakest. austerity and the case for stimulative fiscal policy. For example, fiscal contractions have been shown to correlate with negative As discussed in the next section, these investments can take output outcomes (Blanchard and Leigh 2013), and research various forms, including subsidized or direct job creation, has shown that the positive impact of fiscal stimulus in weak infrastructure investment, or environmental investments. To economies is larger than previously thought. Other work fulfill this role, the federal government should build up a Full (e.g., DeLong and Summers 2012) shows that the existence of Employment Fund (FEF) that can ramp up and down as needed. even minimal, negative long-run impacts of demand shocks can increase the benefits of fiscal stimulus in economies In principle, the FEF could be scaled to the output gap, which, with output gaps, and thus is associated with lower rather as shown in figure 5, persists in recent expansions. More

14 The Importance of Strong Labor Demand realistically, the FEF should be funded like other contingency or DIRECT JOB CREATION emergency programs, meaning it would be treated as mandatory Most economists have little trouble accepting the Federal funding and would not be subject to sequestration or other such Reserve as the lender of last resort when credit markets fail, budget rules. To maximize its effectiveness, the FEF should be as was the case in the financial crisis of 2008. In this section, I triggered on and off by above-average increases in or high levels argue that the persistent absence of full employment in the U.S. of slack variables. Spielberg and I (Bernstein and Spielberg and European labor markets creates a role for the government. 2016) make a similar argument regarding improved triggers This role might not be as an employer of last resort, but the for the extended unemployment insurance benefits program. government should at least engage in some form of direct job There, we argue for triggers based on either levels or changes creation. Surely, the same standard for credit markets should in the underemployment rate (U-6 in the monthly employment apply to the job market: banks facing credit constraints are no report), which includes involuntary part-time workers, making more economically important than the significant numbers of it closer to the Levin measure shown in figure 4. Thus, either workers facing labor demand shortfalls. a high underemployment rate, or one that is rising quickly relative to past values, would trigger FEF outlays. Direct job creation policy exists on a continuum from least to most interventionist. At the less interventionist end are The importance of an automatic trigger for the FEF cannot policies wherein the government subsidizes wages for a set be overstated. If its operations were instead at the discretion period in public or private-sector jobs, including nonprofits. of Congress, political forces would be sure to undermine its Dutta-Gupta et al. (2016) recently completed an exhaustive responsiveness to the business cycle. Though there are many review and evaluation of 40 years of experience with subsidized options for suitable triggers, the underemployment rate is an employment programs. Their review stresses the role of fiscal appealing choice due to its status as a broad measure of labor policy targeting job creation not just during downturns, but market slack. The Bureau of Labor Statistics (BLS) currently during expansions as well: calculates underemployment rates on a quarterly basis at the state level. Sub-state estimates would be much preferable While aggregate labor demand policies—both fiscal and for triggering the FEF, but might be infeasible due to data monetary—are essential to helping low-income workers limitations.5 secure and maintain sufficient employment, additional policies and programs would be valuable throughout the Given the uncertainty in estimating labor market slack, a business cycle for those with serious or multiple barriers relatively small amount of resources—less than $10 billion a to employment. Subsidized employment programs and year—should initially be devoted to the FEF to test the capacity policies are underutilized, potentially powerful tools of the channels noted above and the programs discussed next, for lifting up workers in or at risk of poverty and deep and to gauge the effectiveness of those programs. When the poverty in the United States. These job programs can fund’s trigger turns on—when underemployment either hits a provide income support, an opportunity to engage in trigger level or is quickly rising—FEF funds would be deployed, productive activities, and, in some cases, labor market for example, to support some form of direct job creation. advancement opportunities. They can also offer a In recessions, neither FEF nor any other countercyclical platform for connecting people to other needed services, stimulus spending should be offset with payfors (i.e., tax resources, and networks. [emphasis added] (Dutta- increases or spending cuts used to pay for new spending), Gupta et al. 2016, viii) because these actions would dampen the impact of the Such programs often include a training component; the most stimulus. In expansions, targeted FEF spending should be effective training programs coordinate with local employers to offset with payfors, but Congress must be cautious not to tap ensure that participants are training for in-demand occupations. payfors that hurt one group of vulnerable workers to help a These programs are often directed at particularly disadvantaged different group. Thus, a good way to provide long-term funding workers facing steep barriers to labor market entry associated 6 for the FEF would be a dedicated, progressive tax source. with basic skill deficits, minor physical or cognitive disabilities, Of course, the Federal Reserve must view these dynamics long-term unemployment, discrimination, or criminal records. in the way presented here, recognizing the need for fiscal During the most recent recession, the federal government intervention when aggregate demand is weak. Otherwise, it implemented a successful program from this model through could offset the impact of FEF expenditures and reduce any the Temporary Assistance for Needy Families (TANF) potential demand multiplier effects. A selling point in this emergency fund. As Pavetti (2014) stresses, the TANF program regard is the geographically targeted nature of the FEF. By was really a funding stream to states and localities that could definition, the fund is targeting an area with above-average be used to subsidize employment. She notes that 39 states slack, and should thus be viewed as unlikely to contribute to tapped into the program, using $1.3 billion to place around overheating in the overall economy.

The Hamilton Project • Brookings 15 250,000 low-income people in jobs in less than two years. the nation from investing more than its own savings rates allow. While employers typically received the subsidy for relatively Dean Baker and I point out that this dynamic described the short periods (less than a year), participants often remained demand story in 2000, when the American economy had an in the job market afterward. One careful study from Florida’s unemployment rate of 4 percent and a trade deficit of about that version of the program found that, relative to a control group, same magnitude (Bernstein and Baker 2016). participants’ work and earnings went up not just during the program, but after it as well, suggesting lasting benefits (Roder But in the next expansion the trade deficit’s role was more and Elliott 2013). negative, as an overvalued dollar contributed to a sharp increase in our goods deficit with China (this is the period of the “China At the other, far more interventionist end of the continuum, Shock” documented by Autor, Dorn, and Hanson [2016]), and Paul, Darity, and Hamilton (forthcoming) propose that the the deficit peaked at almost 6 percent of GDP in 2005 and 2006, federal government provide public service jobs for which it pays a historically large imbalance. As Baker and I wrote, “In this salary and benefits. The program creates a National Investment context, the trade deficit was subtracting from demand in the Employment Corps (NIEC) that provides employment grants domestic economy.” to state and local government projects that are “designed to address community needs and provide socially beneficial Thus, it is equally important not to lean too far in the other goods and services to communities and society at large.” direction: trade deficits are not always benign. For one, as Infrastructure, energy efficiency, community development, shown in the first section of this paper we are often not at full education, elder care, art, and other projects could all receive employment, and in periods of weak demand trade deficits are 7 funding through the NIEC. Individuals taking advantage of not being sufficiently offset by other components of growth. the NIEC would have the opportunity for promotions, and Research has shown how some countries attempt to manage Paul, Darity, and Hamilton estimate that the mean salary their savings rates and currencies to maintain trade surpluses, would be about $32,500. They scale their program to eliminate and, since global trade must balance, to impose trade deficits on involuntary unemployment and substantially reduce poverty, other countries. Prominent mainstream economists, including leading to an annual cost of nearly $600 billion, which is close Ben Bernanke (2005, 2015) and Lord Mervyn King (2017), have to what we currently spend on defense. articulated how these imbalances can reduce demand in deficit countries, because surplus countries essentially export excess That is a highly ambitious plan, but as aggregate labor demand savings and import product and labor demand. These impacts has long been insufficient to provide gainful employment on demand become especially important at the aggregate level opportunities to all who seek them, achieving full employment when the economy is at the ZLB. As long as the Fed has ample may well require some degree of direct job creation. Dutta- room to lower interest rates, monetary authorities can help to Gupta et al.’s (2016) review reveals a good track record for well- offset the negative demand impact of the trade deficit. But as the designed programs as well as empirical evidence suggesting risk of encountering the ZLB has gone up, so has the risk that that, once policy helps disconnected workers find their way trade deficits exacerbate the problem of weak aggregate demand. into the labor market, many will try to stay there. From a policy perspective, this analysis suggests two types THE TRADE DEFICIT AND ITS ROLE IN WEAK of interventions. In periods when trade deficits and slack AGGREGATE DEMAND IN THE PRESENCE OF THE coexist, as in the jobless (and initially wage-less) recovery ZERO LOWER BOUND of the 2000s, monetary policy interventions (when the economy has not encountered the ZLB) and fiscal policy In an accounting sense, a trade deficit contributes negatively interventions are effective. This is particularly the case for in the classic, expenditure-side GDP decomposition (GDP fiscal policy responses targeted at places where diminished = private consumption + gross investment + government net exports are clearly taking a toll on employment and spending + exports – imports). However, that simple equation earnings opportunities. In fact, classical trade arguments shows that other GDP components can offset the drag from maintain that whereas trade does create so-called losers (e.g., a trade deficit. Moreover, the trade balance is a function of production workers in richer countries), the gains of trade are exchange rates, relative demand conditions between trading such that winners can compensate losers and still come out partners, trade relations, technologies that affect the logistics ahead. When import competition reduces labor demand in of trade, and more. particular areas, safety net programs, including supply-side and demand-side programs (e.g., training/apprenticeships In periods of truly full employment, trade deficits can expand for a subsidized or guaranteed job), are warranted. These are because a faster-growing economy attracts more imports. In precisely the intended uses of the FEF. that context, imposing balanced trade or even reducing the trade deficit would often be a mistake, because it would prohibit

16 The Importance of Strong Labor Demand At the ZLB or in a global recession, though, it becomes more channels. Failing that, countries can push back against currency important that demand be supported by policies abroad as well manipulation and excess savings through ideas like Bergsten as at home. Lord King calls for a new Bretton Woods (i.e., a and Gagnon’s “countervailing currency intervention” (2012, 1), global agreement for countries to work to move their economies wherein the United States announces “that it would offset the toward balance) that would nudge high-savings countries like effects of currency manipulation through equal purchases of Germany to invest their excess savings more internally, thus the intervening country’s currency. This is intended to deter any reducing capital flows to deficit countries. The U.S. government return of the practice and, like any deterrent if credible, probably should encourage agreements that help ensure sufficient would not have to be used much if at all” (Peterson Institute for demand abroad and clarify their importance via diplomatic International Economics 2017).

The Hamilton Project • Brookings 17 Questions and Concerns

1. Given the increased difficulty of measuring labor market 4. If the Fed increases the rate of inflation, won’t that just slack in recent years, is it possible that slack could be increase nominal wages, but provide no improvement in real overestimated? wages? The fact that policymakers cannot reliably gauge some of the Typically, higher inflation does pass through to higher key metrics in this space, including the natural rate or the nominal wages, which is one reason we should not expect output gap, does not necessarily imply that more-accurate a higher inflation target to hurt workers’ real earnings. The measures would always reveal more slack. In fact, in the late goal of this policy, however, is to avoid the ZLB or insufficient 1970s policymakers overestimated potential GDP, which led aggregate demand more generally, where too many workers to high and damaging levels of inflation and unemployment. are unemployed and face stagnant real wage growth. These relationships and these variables are dynamic, and economists must allow for biases in both directions. 5. Won’t your proposals for more aggressive fiscal policy cause larger deficits and debt levels? 2. Is it likely that the FEF and direct jobs creation programs I am clearly calling for more spending both during downturns— will be effectively implemented? through more responsive automatic stabilizers—and during Ideas like the FEF or direct jobs creation depend on a expansions (“sustained fiscal policy”). These proposals need functional government sector that can efficiently implement not have a large impact on long-run deficits or debt levels. such programs. For example, if, under a direct job creation First, Congress already provides discretionary fiscal support program, employers simply substitute subsidized for during most recessions. The goal of one of my proposals— nonsubsidized workers, there is no addition to aggregate an FEF that is triggered by need—is to ensure that the fiscal demand. For this reason, it is always a good idea to try new impetus is both timely and well-designed rather than delayed programs on a pilot basis before taking them national. or distorted by extended periods of political bargaining. In addition, by shrinking the amount of time the economy is 3. Slack in the U.S. labor market appears to be very limited in below full employment, the policies should both boost revenue early 2018. Does this undermine the case for your proposals? and lift the denominator (GDP) in debt-to-GDP ratios. Finally, It is true that the U.S. economy in early 2018 is quite clearly I recommend raising more revenues as needed, preferably closing in on full employment; the unemployment rate is well through progressive tax policies. below the natural rate as estimated by various agencies like the 6. You suggest adding non-experts in monetary policy— CBO and the Federal Reserve Board, though the absence of specifically, advocates for low-wage and minority workers— wage and price pressures suggests that we have not yet reached to your proposed process led by the Federal Reserve that full capacity. This may lead some readers to question whether would evaluate and revise the monetary policy framework. we have already solved the aggregate demand problem! Of Won’t that both slow the process and make it needlessly more course, that would be a mistake. My point is not that the U.S. contentious? economy never achieves sufficient levels of demand: it is that periods like the present are too infrequent, and policymakers It may or may not have these impacts. Progressive groups like need an aggressive agenda to implement when labor market “Fed Up,” while critical of some Fed actions, have developed slack is much greater than it is today. good and useful relationships with the central bank. But such additions are absolutely essential for broad public agreement about the outcome of the framework evaluation process. Moreover, the workers represented by these groups are often the ones most affected, for better or worse, by Fed policies, and they therefore very much deserve to play a role in shaping those policies.

18 The Importance of Strong Labor Demand Conclusion

mong the many assumptions made by economists, beyond recession fighting to encompass sustained fiscal policy one of the most empirically indefensible is that the during weak expansions. I therefore propose a mandatory U.S. economy is generally at full employment. It is also Full Employment Fund that expands and contracts with need. Aan assumption with the capacity to do tremendous damage to Third, as one use for this fund, I propose measures providing people and communities who, because of inadequate demand for direct job creation. Finally, because persistent trade deficits and thus limited economic opportunities, face stagnating in the presence of the ZLB can constitute a drag on demand, living standards. Conventional measures reveal persistent I propose policies to both restore lost demand and reduce the slack in recent decades, and this slack disproportionately hurts trade deficits themselves. those with the fewest economic resources. These four responses represent a small start in addressing I have suggested a four-part policy response. First, given the this critically important market failure. Much more research diminished correlation between unemployment and inflation, is needed to identify the extent of weak aggregate demand. along with the increased risk of hitting the zero lower We must improve our measurement of output gaps and labor bound (ZLB) on the federal funds rate, the monetary policy market slack, investigate the factors explaining the absence of framework should be reformed to reduce the risk of hitting full employment, explore geographical variation in slack, and the ZLB and ensure that the central bank has the ability to examine other policies that can play a role in explaining labor support the economy during a downturn. Second, we must market slack. But the first step is recognizing the problem and expand our thinking about fiscal policy and aggregate demand working toward its solution.

The Hamilton Project • Brookings 19 Author

Jared Bernstein Senior Fellow, Center on Budget and Policy Priorities

Jared Bernstein joined the Center on Budget and Policy He is the author and coauthor of numerous books for both Priorities in May 2011 as a Senior Fellow. From 2009 to 2011, popular and academic audiences, including Getting Back to Full Bernstein was the Chief Economist and Economic Adviser Employment: A Better Bargain for Working People, Crunch: Why to Vice President , Executive Director of the White Do I Feel So Squeezed?, nine editions of The State of Working House Task Force on the Middle Class, and a member of America, and his latest book, The Reconnection Agenda: Reuniting President Obama’s economic team. Growth and Prosperity. Bernstein has published extensively in various venues, including , The Washington Bernstein’s areas of expertise include federal and state economic Post, and the Financial Times. He is an on-air commentator and fiscal policies, income inequality and mobility, trends in for the cable stations CNBC and MSNBC, contributes to The employment and earnings, international comparisons, and Washington Post’s PostEverything blog, and hosts On The the analysis of financial and housing markets. Economy (jaredbernsteinblog.com).

Prior to joining the Obama administration, Bernstein was Bernstein holds a PhD in Social Welfare from Columbia a senior economist and the director of the Living Standards University. Program at the Economic Policy Institute in Washington, D.C.

Between 1995 and 1996, he held the post of Deputy Chief Economist at the U.S. Department of Labor.

Acknowledgments

The author thanks Somin Park and Ben Spielberg for research assistance, and Richard Kogan, Becca Portman, Jay Shambaugh, and Hamilton Project authors’ conference participants for helpful suggestions.

20 The Importance of Strong Labor Demand Endnotes

1. Using an econometric technique developed by Blanchard and Quah (1989) to separately identify supply and demand shocks, they derive the potential GDP series shown in the figure. By the end of their data, while CBO’s current estimate of potential GDP is coincident with the actual value, Coibion, Gorodnichenko, and Ulate’s (2017) measure is 11 percent, or about $2 trillion higher. Interestingly, that is about the same difference between CBO’s 2007 prediction of potential GDP today and the most recent estimate for 2017Q2. Ball et al. (2014) come to a similar conclusion as Coibion, Gorodnichenko, and Ulate. 2. The estimates come from fixed effects panel regressions for the period from 1979 to 2015 that regress the log real hourly wage on the slack measure, where the slack measures are logged and lagged one year. 3. The model is run with data from 1968 to 2016 and regresses the percentage change in nominal median household income on lagged inflation (CPI-U- RS), the percentage change in wages, employment, and the unemployment rate gap. The R-squared in such a regression is 0.8. 4. This idea is somewhat like the Bank of Canada’s five-year reviews of its monetary policy framework, though I am suggesting a process that is considerably more inclusive than the Bank of Canada’s (as I understand it), and involves no direct government involvement. 5. However, BLS often uses modeling procedures to develop sub-state estimates (e.g., in the BLS Local Area Unemployment Statistics program), which could be applicable here as well. 6. One option is a small financial transaction tax, as other authors and I have described (e.g., Bernstein 2015, 2016; Burman et al. 2015). 7. Deficits at the sectoral level may be important as well, separate from their implications for aggregate demand. If production is concentrated geographically (as with some types of manufacturing) deficits can have important impacts at the community level.

The Hamilton Project • Brookings 21 References

Auerbach, Alan J., and Yuriy Gorodnichenko. 2017, September. Binder, Carola, and Alex Rodrigue. 2016, September. “Monetary “Fiscal Stimulus and Fiscal Sustainability.” Working Rules and Targets: Finding the Best Path to Full Paper 23789, National Bureau of Economic Research, Employment.” Policy Futures, Center on Budget and Policy Cambridge, MA. Priorities, Washington, DC. Autor, David H., David Dorn, and Gordon H. Hanson. 2016, Blanchard, Olivier, and Daniel Leigh. 2013, February. “Growth January. “The China Shock: Learning from Labor Market Forecast Errors and Fiscal Multipliers.” Working Paper Adjustment to Large Changes in Trade.” Working Paper 18779, National Bureau of Economic Research, Cambridge, 21906, National Bureau of Economic Research, Cambridge, MA. MA. Blanchard, Olivier Jean, and Danny Quah. 1989. “The Dynamic Ball, Laurence M. 2014, May. “Long-term Damage from the Great Effects of Aggregate Demand and Supply Disturbances.” Recession in OECD Countries.” Working Paper 20185, American Economic Review 79 (4): 655–73. National Bureau of Economic Research, Cambridge, MA. Blanchard, Olivier, and Lawrence Summers. 2017, October. Ball, Laurence, Brad DeLong, and Larry Summers. 2014, April. “Rethinking Stabilization Policy: Back to the Future.” “Fiscal Policy and Full Employment.” Full Employment Peterson Institute for International Economics, Project, Center on Budget and Policy Priorities, Washington, DC. Washington, DC. Brookings Institution. 2018, January. “Should the Fed Stick with Bergsten, C. Fred, and Joseph E. Gagnon. 2012, December. the 2 Percent Inflation Target or Rethink It?” Past Event, “Currency Manipulation, the US Economy, and the Global Brookings Institution, Washington, DC. Economic Order.” Policy Brief PB12-25, Peterson Institute Bureau of Economic Analysis (BEA). 2007–17. “Gross Domestic for International Economics, Washington, DC. Product: Implicit Price Deflator [GDPDEF].” FRED Bernanke, Ben. 2005, April 14. “Remarks by Governor Ben S. Economic Data, Federal Reserve Bank of St. Louis, St. Bernanke.” Presented at the Sandridge Lecture, Virginia Louis, MO. Association of Economists, Richmond, VA. ———. 2017. “Personal Consumption Expenditures Excluding ———. 2013, February 26. “Semiannual Monetary Policy Report Food and Energy (Chain-Type Price Index) [PCEPILFE].” to the Congress.” Testimony before the Committee FRED Economic Data, Federal Reserve Bank of St. Louis, on Banking Housing, and Urban Affairs, U.S. Senate, St. Louis, MO. Washington, DC. Bureau of Labor Statistics (BLS). 1949–2017. “Current Population ———. 2015, April. “Why Are Interest Rates So Low, Part 3: The Survey.” Bureau of Labor Statistics, U.S. Department of Global Savings Glut.” Brookings Institution, Washington, Labor, Washington, DC. DC (blog). Burman, Leonard E., William G. Gale, Sarah Gault, Bryan Kim, ———. 2017, October. “Temporary Price-level Targeting: An James R. Nunns, and Steven M. Rosenthal. 2015, June. Alternative Framework for Monetary Policy.” Brookings “Financial Transaction Taxes in Theory and Practice.” Tax Institution, Washington, DC (blog). Policy Center, Urban Institute & Brookings Institution, Bernstein, Jared. 2015, July 22. “The Case for a Tax on Financial Washington, DC. Transactions.” New York Times. Coibion, Olivier, Yuriy Gorodnichenko, and Mauricio Ulate. 2017, ———. 2016, June 13. “We’re Going to Need More Tax Revenue. October. “The Cyclical Sensitivity in Estimates of Potential Here’s How to Raise It.” American Prospect Longform. Output.” Working Paper 23580, National Bureau of Bernstein, Jared, and Dean Baker. 2016, December 8. “Why Trade Economic Research, Cambridge, MA. Deficits Matter.” Atlantic. Congressional Budget Office (CBO). 2007, January. “The Budget Bernstein, Jared, and Ben Spielberg. 2016, March. “Preparing for the and Economic Outlook: Fiscal Years 2008 to 2017.” Next Recession: Lessons from the American Recovery and Congressional Budget Office, U.S. Congress, Washington, Reinvestment Act.” Center on Budget and Policy Priorities, DC. Washington, DC. ———. 2017, June. “An Update to the Budget and Economic Bernstein, Jared, Ben Spielberg, and Keith Bentele. Forthcoming. Outlook: 2017 to 2027.” Congressional Budget Office, U.S. “The Relationship between Tight Labor Markets and the Congress, Washington, DC. Earnings of Low-Income Households.”

22 The Importance of Strong Labor Demand Council of Economic Advisers. 2016, February. “Chapter 2: The Year Levin, Andrew. 2014. “The Design and Communication of in Review and the Years Ahead.” In 2016 Economic Report of Systematic Monetary Policy Strategies.” Journal of the President. Washington, DC: U.S. Government Printing Economic Dynamics and Control 49 (C): 52–69. Office. Paul, Mark, William Darity, Jr., and Darrick Hamilton. “The DeLong, J. Bradford, and Lawrence H. Summers. 2012. “Fiscal Federal Job Guarantee: A Policy to Achieve Full Policy in a Depressed Economy.” Brookings Papers on Employment.” Full Employment Project, Center on Budget Economic Activity (Spring): 233–97. and Policy Priorities, Washington, DC. DeLong, Brad, Larry Summers, and Laurence Ball. 2014, April. Pavetti, LaDonna. 2014. “Subsidized Jobs: Providing Paid “Fiscal Policy and Full Employment.” Policy Futures, Employment Opportunities When the Labor Market Fails.” Center on Budget and Policy Priorities, Washington, DC. Policy Futures, Center on Budget and Policy Priorities, Dutta-Gupta, Indivar, Kali Grant, Matthew Eckel, and Peter Washington, DC. Edelman. 2016. Lessons Learned from 40 Years of Subsidized Peterson Institute for International Economics. 2017, June Employment Programs: A Framework, Review of Models, 6. “Peterson Institute’s Study by Bergsten and and Recommendations for Helping Disadvantaged Workers. Gagnon Proposes New Strategy to Counter Currency Washington, DC: Georgetown Center on Poverty and Manipulation.” Press Release, Peterson Institute for Inequality. International Economics, Washington, DC. Economic Policy Institute. 2017. “State of Working America Data Robb, Greg. 2017, May. “Fed’s Williams Backs Changing Central Library: Wages by Percentile.” Economic Policy Institute, Bank’s Strategy to Price-level Targeting.” MarketWatch. Washington, DC. Roder, Anne, and Mark Elliott. “Stimulating Opportunity: An Federal Open Market Committee (FOMC). 2017, September. Evaluation of ARRA-Funded Subsidized Employment “Economic Projections of Federal Reserve Board Members Programs.” Economic Mobility Corporation, New York, and Federal Reserve Bank Presidents under Their NY. Individual Assessments of Projected Appropriate Monetary Shambaugh, Jay. 2012. “The Euro’s Three Crises.” Brookings Papers Policy, September 2017.” Released 2:00 p.m., EDT, on Economic Activity (Spring): 157–231. September 20, 2017. Federal Open Market Committee, Summers, Lawrence H. 2014. “U.S. Economic Prospects: Secular Washington, DC. Stagnation, Hysteresis, and the Zero Lower Bound.” ———. n.d. “Summary of Economic Projections.” Federal Open Business Economics 49 (2): 65–73. Market Committee, Washington, DC. Svensson, Lars E. O. 2017, August. “Leaning Against the Wind: The Furman, Jason. 2016, October. “The New View of Fiscal Policy Role of Different Assumptions About the Costs.” Working and Its Application.” Conference: Global Implications Paper 23745, National Bureau of Economic Research, of Europe’s Redesign. Center on Global Economic Cambridge, MA. Governance, School of International and Public Affairs, Williams, John C. 2017, February. “Three Questions on R-star.” , New York, NY. FRBSF Economic Letter, Federal Reserve Bank of San Glassman, Jim. 2017, June. “Will the Fed Raise Its Inflation Target?” Francisco, San Francisco, CA. Commercial Banking, J.P. Morgan Chase, New York, NY. Yagan, Danny. 2017, September. “Employment Hysteresis from the Harrison, David. 2017, April 2. “Rethinking the Widely Held 2% Great Recession.” Working Paper 23844, National Bureau Inflation Target.” Wall Street Journal. of Economic Research, Cambridge, MA. Irwin, Neil. 2016, February 12. “Negative 0.5% Interest Rate: Why Yellen, Janet L. 2014, July 2. “Monetary Policy and Financial People Are Paying to Save.” New York Times. Stability.” Speech given at the 2014 Michel Camdessus King, Mervyn. 2017, May. “World Trade and Exchange Rates: From Central Banking Lecture, International Monetary Fund, the Pax Americana to a Multilateral New Order.” Peterson Washington, DC. Board of Governors of the Federal Institute for International Economics, Washington, DC. Reserve System, Washington, DC. Kogan, Richard, Chad Stone, Bryann DaSilva, and Jan Rejeski. 2015, February. “Difference Between Economic Growth Rates and Treasury Interest Rates Significantly Affects Long-Term Budget Outlook.” Center on Budget and Policy Priorities, Washington, DC.

The Hamilton Project • Brookings 23 ADVISORY COUNCIL

GEORGE A. AKERLOF RICHARD GEPHARDT PENNY PRITZKER University Professor President & Chief Executive Officer Chairman Georgetown University Gephardt Group Government Affairs PSP Partners

ROGER C. ALTMAN ROBERT GREENSTEIN MEEGHAN PRUNTY Founder & Senior Chairman Founder & President Managing Director Evercore Center on Budget and Policy Priorities Blue Meridian Partners Edna McConnell Clark Foundation KAREN ANDERSON MICHAEL GREENSTONE Senior Director of Policy and Communications Milton Friedman Professor of Economics ROBERT D. REISCHAUER Becker Friedman Institute for Director of the Becker Friedman Institute for Distinguished Institute Fellow & President Emeritus Research in Economics Research in Economics Urban Institute The University of Chicago Director of the Energy Policy Institute University of Chicago ALICE M. RIVLIN ALAN S. BLINDER Senior Fellow, Economic Studies Gordon S. Rentschler Memorial Professor of GLENN H. HUTCHINS Center for Health Policy Economics & Public Affairs Co-founder The Brookings Institution Princeton University North Island Nonresident Senior Fellow Co-founder DAVID M. RUBENSTEIN The Brookings Institution Silver Lake Co-Founder & Co-Chief Executive Officer ROBERT CUMBY JAMES A. JOHNSON The Carlyle Group Professor of Economics Chairman Georgetown University Johnson Capital Partners ROBERT E. RUBIN Former U.S. Treasury Secretary STEVEN A. DENNING LAWRENCE F. KATZ Co-Chair Emeritus Chairman Elisabeth Allison Professor of Economics Council on Foreign Relations General Atlantic Harvard University LESLIE B. SAMUELS JOHN M. DEUTCH MELISSA S. KEARNEY Senior Counsel Institute Professor Professor of Economics Cleary Gottlieb Steen & Hamilton LLP Massachusetts Institute of Technology University of Maryland Nonresident Senior Fellow SHERYL SANDBERG CHRISTOPHER EDLEY, JR. The Brookings Institution Chief Operating Officer Co-President and Co-Founder Facebook The Opportunity Institute LILI LYNTON Founding Partner DIANE WHITMORE SCHANZENBACH BLAIR W. EFFRON Boulud Restaurant Group Margaret Walker Alexander Professor Partner Director Centerview Partners LLC HOWARD S. MARKS The Institute for Policy Research Co-Chairman Northwestern University DOUGLAS W. ELMENDORF Oaktree Capital Management, L.P. Nonresident Senior Fellow Dean & Don K. Price Professor The Brookings Institution of Public Policy MARK MCKINNON Harvard Kennedy School Former Advisor to George W. Bush RALPH L. SCHLOSSTEIN Co-Founder, No Labels President & Chief Executive Officer JUDY FEDER Evercore Professor & Former Dean ERIC MINDICH McCourt School of Public Policy Chief Executive Officer & Founder ERIC SCHMIDT Georgetown University Eton Park Capital Management Technical Advisor Alphabet Inc. ROLAND FRYER ALEX NAVAB Henry Lee Professor of Economics Former Head of Americas Private Equity ERIC SCHWARTZ Harvard University KKR Chairman and CEO Founder 76 West Holdings JASON FURMAN Navab Holdings Professor of the Practice of THOMAS F. STEYER Economic Policy SUZANNE NORA JOHNSON Business Leader and Philanthropist Harvard Kennedy School Former Vice Chairman Senior Counselor Goldman Sachs Group, Inc. LAWRENCE H. SUMMERS The Hamilton Project Charles W. Eliot University Professor PETER ORSZAG Harvard University MARK T. GALLOGLY Vice Chairman of Investment Banking Cofounder & Managing Principal Managing Director and LAURA D’ANDREA TYSON Centerbridge Partners Global Co-head of Health Professor of Business Administration and Lazard Economics Director TED GAYER Nonresident Senior Fellow Institute for Business & Social Impact Vice President & Director The Brookings Institution Berkeley-Haas School of Business Economic Studies The Brookings Institution RICHARD PERRY Managing Partner & JAY SHAMBAUGH TIMOTHY F. GEITHNER Chief Executive Officer Director President Perry Capital Warburg Pincus

24 The Importance of Strong Labor Demand ADVISORY COUNCIL Highlights

By conventional measures, the U.S. job market has suffered some degree of slack for about 70 percent of the time since 1980. The absence of persistent, strong labor market demand has a significant negative impact on wages and incomes, with these costs falling disproportionately on the least advantaged. In this paper, Jared Bernstein offers a four-part proposal to increase labor demand along with earnings and employment opportunities.

The Proposal

Reform the monetary policy framework to accommodate more monetary stimulus and reduce the risk of hitting the zero lower bound on interest rates.

Develop a Full Employment Fund that automatically expands and contracts with changes in the business cycle to provide fiscal stimulus when and where it is needed.

Support direct job creation programs to subsidize employment and help disconnected workers to enter the labor market.

Design international trade policies to safeguard aggregate demand and mitigate the negative effects of trade deficits.

Benefits

These four policy proposals would begin to address the problem of persistent labor market slack. When labor market slack is significantly reduced, workers can more easily find employment, move across firms, and achieve career progress. Implementing these policies would boost both wages and overall economic growth.

1775 Massachusetts Ave., NW Washington, DC 20036

(202) 797-6484

Printed on recycled paper. WWW.HAMILTONPROJECT.ORG The Hamilton Project • Brookings 25